- Email: [email protected]
Phosphorylation of protein components of isolated zymogen granule membranes from the rat pancreas
Volume
49, number
2
FEBS LETTERS
PHOSPHORYLATION
OF PROTEIN COMPONENTS OF ISOLATED MEMBRANES FROM THE RAT PANCREAS Monique LAMBERT,.Jean
Department
December
of Biochemistry,
ZYMOGEN
1974
GRANULE
CAMUS and Jean CHRISTOPHE
Brussels University School of Medicine, 1000 Brussels, Waterloo Boulevard 11.5, Belgium
Received
5 November
1. Introduction Our earlier results have already documented some aspects of protein phosphorylation in rat pancreas fragments [l] . The stimulation of amylase secretion by pancreozymin and caerulein was accompanied in vitro by an increase in 32P orthophosphate incorporation into proteins, and especially so in the membrane proteins of zymogen granules. The mechanism involved in the activation by pancreozymin of protein kinase(s) directing these phosphorylations is not clearly established The role of cyclic AMP as an intracellular mediator is controversial [2-41 and recent evidence supports the participation of cyclic GMP [4]. The aim of the present investigation was to examine some properties of the isolated membranes of zymogen granules from the rat pancreas. We were able to detect an endogenous protein kinase activity capable of phosphorylating 8 to 10 membrane protein constituents as well as added histones. Cyclic AMP, and cyclic GMP to a lesser extent, exerted moderate stimulatory effects. These effects of cyclic nucleotides on membrane phosphorylation became more apparent upon addition of a partially purified soluble protein kinase also extracted from the rat pancreas.
1974
A crude protein kinase was also prepared. Rat pancreases were homogenized in 5 mM phosphate buffer (pH 7.0) containing 4 mM EDTA and 75 mM e-aminocaproate. The homogenate was centrifuged at 27 000 X g for 30 min. Ammonium sulfate was added to the supernatant to a final concentration of 30%. The precipitate was collected, dissolved in 5 mM phosphate (pH 7.0), 2 mM EDTA, 75 mM e-aminocaproate, and dialyzed overnight against the same buffer. This preparation was kept at -90°C. The basal assay mixture for studying membrane phosphorylation contained, in a total volume of 0.2 ml, 200 pg membrane proteins (as determined by the Folin assay), 1O-20 PM [y-3 * P] ATP (1 to 4.1 O3 cmp/pmole; the Radiochemical Centre, Amersham, England), 50 mM Tris-HCl (pH 7.5), 1 mM aminophylline, Trasylol (Bayer, Leverkussen, Germany) 500 UIK/ml, and 10 mM MgS04. When specified, the medium was supplemented with 10 mM NaF, histones II A (Sigma, St. Louis, Missouri), 10e6 M cyclic or 5.10” M cyclic GMP. The phosphorylation was initiated by the addition of radioactive ATP and conducted at 30°C.
3. Results and discussion 2. Materials and methods Zymogen granules were isolated from rat pancreas as described previously [ 11. Secretory proteins were released in 0.2 M bicarbonate at pH 8.2. Membranes were purified by discontinuous gradient centrifugation according to Meldolesi [S] . They were utilized whitin the same day for the phosphorylation tests. 228
Fig. 1 shows that phosphorylation proceeded rapidly and that after 60 set a plateau was maintained for at least 5 min. This reaction was not limited by the hydrolysis of [r-32P] ATP by membranous ATPase. Indeed, the addition of an ATP-regenerating system made of 7.5 mM phosphoenolpyruvate and 25 U/ml pyruvate kinase did not improve the kinetics (data not shown). On the other hand, such a membrane preparatior North-Holland Publishing Company - Amsterdam
FEBS LETTERS
Volume 49, number 2 ,100
75
__.-. ____/---~
.:"f._._
0
___.__-.-.x~ "----,.-;k-~_____-______ _n
I
I
I
1
2
3
--------_____;
time
I
4
(min)
n
1 5
Fig. 1. Time-study of autophosphorylation of the membrane of zymogen granules with [T- ‘sP] ATP. This experiment was conducted as described under Materials and methods and is representative of 3 similar experiments. The phosphorylations were stopped by the addition of 1% Ccl, COOH and the pellets were filtered on Millipore AAWP (Bedford, U.S.A.). basal Results of one representative experiment (n=3). ( e) medium; (x-.x-.x) 10 mM NaF; ( o-c-o) l-lo6 M cyclic AMP.
had a moderate phosphoprotein phosphatase activity which was inhibited by 50% by 10 mM NaF (P. Robberecht, personal communication). Fig. 1 illustrates that in the presence of 10 mM NaF, protein phosphorylation exceeded control values. The plateau appears therefore to result from an equilibrium between the opposite activities of protein kinase and phosphoprotein phosphatase in the zymogen granule membrane. l.10e6 M cyclic AMP induced a moderate but consistent stimulation of autophosphorylation.
December 1974
Table 1 shows a comparison of the endogenous activity of membranes with a preparation of soluble protein kinase(s) partially purified from the rat pancreas. The autophosphorylation of membranes was only slightly stimulated by 1.10e6 M cyclic AMP or 5a104 M cyclic GMP and the protein kinase activity exerted by these membranes on added histones II A was unaffected by cyclic nucleotides. On the other hand, the activity of soluble protein kinase(s) was markedly stimulated by 1.10” M’cyclic AMP and 5.10’ M cyclic GMP. When this soluble preparation was added to membranes, the aggregate phosphorylation was lower than the sum of each autophosphorylation activities considered separately. The marked effect of cyclic nucleotides on this mixture suggests that the activity of soluble protein kinase(s) was predominating and contributed to the phosphorylation of membrane proteins. It can be hypothesized that a component of the crude soluble protein kinase preparation inhibited membrane protein kinase activity. It is more likely however that the affinity of semi-purified soluble protein kinase(s) for endogenous substrates was lower than that for membrane proteins. Competitive inhibitio might render the autophosphorylation of soluble protein negligible in the presence of zymogen granule membranes. As shown in Table 2, extraction of membranes with 0.5 M KC1 did not decrease the protein kinase activity measured on endogenous substrates or on added histones. In addition, this treatment did not affect the capacity of membranes to be phosphorylated in the
Table 1 Comparison of endogenous and exogenous protein kinase activities Proteins added
pmoles phosphate/mg
Soluble Histones no cyclic Zymogen nucleotide granule protein II A added membrane kinase (250 rg) (250 clg) (158 I.& + 0 0 47 0 + 0 21 + + 0 27 ___--__--___________-----------------___---___ + 0 + 125 0 + + 531
protein. 10 min
cyclic AMP
cyclic GMP
1.1O-6 M 58 33 59
5.10-6 M 52 26 44
125 1004
116 811
Assays were performed in the presence of 10 mM NaF. Results of one representative experiment (n=4).
229
FEBS LETTERS
Volume 49, number 2
Table 2 Protein kinase activity of the membrane of zymogen granules after treatment with 0.5 M KC1 Proteins added
No addition Soluble protein kinase (250 pg) Histones II A (250 fig)
pmoles phosphate/mg protein. 5 min Control membranes
Washed membranes
48 110 64
73 129 66
Purified membrane pellets were suspended at 0°C in 50 mM Tris-HCl (pH 7.5) and Trasylol500 UIK/ml in the presence or absence (control) of 0.5 M KCl. These suspensions were recentrifuged for 10 min at 2500 X g. The recovered membranes were then utilized for phosphorylation tests in the presence of 10 mM NaF. Results of one respresentative experiment (n=3).
cm
December 1974
presence of exogenous protein kmase(s). Thus it appears that protein kinase activity and protein substrates for this enzyme were intrinsic parts of the zymogen membrane and can not be assimilated to exogenous proteins simply contaminating the membrane. Individual’protein components of the zymogen granule membrane were analyzed by polyacrylamide gel electrophoresis in sodium dodecylsulfate. Staining with Coomassie blue revealed at least 26 distinct bands (fig. 2A). The crude soluble protein kinase preparation when submitted to similar electrophoresis showed 10 to 12 bands, the main component having a mol.wt of 66 000 (fig. 2B). As can be seen in fig. 3,8 to 10 components of
moved
Fig. 2. Polyacrylamide gel electrophoresis in 0.1% sodium dodecylsulfate after 10 min incubation of: A: zymogen granule membrane from rat pancreas (a indicates the protein band most markedly phosphorylated in the high molecular weight range, as illustrated in fig. 3). B: semi-purified soluble protein kinase preparation from rat pancreas. C: a mixture of the two preceding extracts. The standard incubation medium described under Materials and methods was suplemented with 10 mM NaF. Each incubation was stopped by the addition of 0.1 ml 0.03 M phosphate buffer (pH 7.2), 3% (v/v) fl-mercaptoethanol, 24% sucrose, 6% sodium dodecylsulfate. These samples were heated for 2 min at 100°C. 25 ~1 samples were then applied to polyacrylamide gels containing 10% acrylamide, 0.1% sodium dodecylsulfate, 0.1 M sodium phosphate (pH 7.2), 0.125% persulfate, and 0.0135% Temed (from Fluka, Buchs, Switzerland). Electrophoresis was carried out for 270 min at 8 mA/tube. These gels were stained with Coomassie blue in 10% acetic acid-25% isopropanol and destained in 10% acetic acid-lo% isopropanol.
230
December 1974
FEBS LETTERS
Volume 49, number 2 c
w llnOUT 600
AMP
w Ill4
ACTIVITY
BASAL
CAMP
lq6 M
BASAL
ACTIVITY
500
coo
l-l
-100
300 U300 ” _ 100 v) 0
Iso
E
-40
0
-30 -20
u 3
-10
s
I* Y 0
3 ” w -8 600
ii-
0 *
0 s
KINASE
PROTEIN
K INASE
PROTEIN
>
I
ADDED
D
500
Y I-
4
z ;
p 400 0.
100
CJ300 200
loo
\
-\
0_1
0
I
IO
20
SLICE
I
30
‘\ I
LO
!
50
No
r
0
,
I
to
20 SLICE
I
LO
30
I
z
I
:8
0 0
30 20
2
IO
50
No
Fig. 3. Distribution of aZP among the proteins of phosphorylated zymogen granule membranes, after separation by polyacrylamide gel electrophoresis in 0.1% sodium dodecylsulfate. After 10 min incubation carried out under the same general conditions as those in fig. 2, in the presence of 10 mM NaF and with or without 1=106 M cyclic AMP, 100 nl of SDS-solubilized proteins were applied to each gel. After electrophoresis, the gels were fixed overnight in 10% acetic acid-25% isopropanol, washed in 10% acetic acid-10% isopropanol and cut into 2 mm slices. The radioactivity was.measured by scintillation counting. The autophosphorylation profile of the isolated crude soluble protein kinase preparation is illustrated by broken lines in the two lower panels. The linear relationship between molecular weight and the distance of migration was standardized with bovine serum albumin, hen ovalbumin, bovine carboxypeptidase A, bovine ribonuclease and horse heart cytochrome c.
the membrane with mol.wt ranging from 12 000 to 70 000 were phosphorylated by endogenous kinase. There was no specific effect of cyclic AMP on a particular protein species. These results are at variance with those reported by MacDonald and Ronzio [6] while our work was in progress. These authors noticed that more than 70% of 32P was incorporated into a single protein of approximate mol.wt 130 000 in zymogen granule membranes incubated for 5 set
without aminophylline and NaF (this component was isolated by SDS polycrylamide gel electrophoresis at pH 2.4). After incubation in the presence of the exogenous kinase, we detected 8 phosphorylated components. Cyclic AMP increased the incorporation of y-32P in all molecular species but its effect was particularly pronounced for a protein with a mol.wt of approx. 70 000 (Band a in fig. 2A). In conclusion, it appears that 8 to 10 proteins in 231
Volume 49, number 2
FEBS LETTERS
the zymogen granule membrane can be phosphorylated by an endogenous protein kinase. In addition, soluble protein kinase(s) can also phosphorylate the membrane of zymogen granules. The endogenous membrane protein kinase activity was only slightly dependent on cyclic nucleotides. If we assume that this kinase remained unaltered during membrane isolation, this observation might suggest that the in vivo phosphorylation of zymogen granule membrane is not under hormonal control. This conclusion would contradict our previous hypothesis, based on intact cell preparation [l] , according to which phosphorylation of specific proteins could change granule membrane properties and facilitate the fusion with the plasma membrane during exocytosis. It is however possible to reconcile the data obtained on pancreas fragments and pancreas extracts if we admit that rises in the intracellular levels of cyclic GMP or cyclic AMP entail a dissociation of cytoplasmic protein kinase(s) am’ that the liberated catalytic subunit(s) are secondarily transferred on zymogen granule membranes, resulting in ‘in situ’ phosphorylation of 8 to 10 constituent proteins. This hypothesis finds some support __ in recent observations. Korenman et al. [7] demonstrated a translocation of cytoplasmic protein kinase towards microsomes, in a cyclic AMP-independent form, after stimulating rat uterus with isoproterenol. Our own preliminary results (unpublished data) indicate an important rise in membrane bound protein kinase activity when zymogen granule membranes of
232
December 1974
rat pancreas were incubated at O’C in the presence of semi-purified soluble protein kinase and of 1 *lo” M cyclic AMP.
Acknowledgements This work was supported by Fonds de la Recherche Scientifique MBdicale (Belgium), grant No. 20 403. We wish to express our thanks to Dr Wald (Bayer, Brussels), for the gift of Trasylol and to Mrs Ballinckx and Mr Gelston for preparation of the manuscript.
References [l] Lambert, M., Camus, I. and Christophe, J. (1973) Biochem.
Biophys. Res. Commun. 52,935-942. [2] Case, R. M., Johnson, M., Scratcherd, T. and Sherratt, H. S. A (1972) J. Physiol. 223,669-684. [31 Rutten, W. J., de Pont, J. J. H. H. M. and Bonting, S. L. (1972) Biochim. Biophys. Acta 274, 201-213. [41 Robberecht, P., Deschodt-Lanckman, M., de Neef, P., Borgeat, P. and Christophe, J. (1974) FEBS Lett. 43. 139-143. ISI Meldolesi, J., Jamieson, J. D., Palade, G. E. and Evans, L. M (1971) J. Cell. Biol. 49, 109-129. I61 MacDonald, R. J. and Ronzio, R. A. (1974) FEBS Lett. 40,203-206. [71 Korenman, S. G., Bhalla, R. C., Sanborn, B. M. and Stevens, R. H. (1974) Science 183,430-432.
49, number
2
FEBS LETTERS
PHOSPHORYLATION
OF PROTEIN COMPONENTS OF ISOLATED MEMBRANES FROM THE RAT PANCREAS Monique LAMBERT,.Jean
Department
December
of Biochemistry,
ZYMOGEN
1974
GRANULE
CAMUS and Jean CHRISTOPHE
Brussels University School of Medicine, 1000 Brussels, Waterloo Boulevard 11.5, Belgium
Received
5 November
1. Introduction Our earlier results have already documented some aspects of protein phosphorylation in rat pancreas fragments [l] . The stimulation of amylase secretion by pancreozymin and caerulein was accompanied in vitro by an increase in 32P orthophosphate incorporation into proteins, and especially so in the membrane proteins of zymogen granules. The mechanism involved in the activation by pancreozymin of protein kinase(s) directing these phosphorylations is not clearly established The role of cyclic AMP as an intracellular mediator is controversial [2-41 and recent evidence supports the participation of cyclic GMP [4]. The aim of the present investigation was to examine some properties of the isolated membranes of zymogen granules from the rat pancreas. We were able to detect an endogenous protein kinase activity capable of phosphorylating 8 to 10 membrane protein constituents as well as added histones. Cyclic AMP, and cyclic GMP to a lesser extent, exerted moderate stimulatory effects. These effects of cyclic nucleotides on membrane phosphorylation became more apparent upon addition of a partially purified soluble protein kinase also extracted from the rat pancreas.
1974
A crude protein kinase was also prepared. Rat pancreases were homogenized in 5 mM phosphate buffer (pH 7.0) containing 4 mM EDTA and 75 mM e-aminocaproate. The homogenate was centrifuged at 27 000 X g for 30 min. Ammonium sulfate was added to the supernatant to a final concentration of 30%. The precipitate was collected, dissolved in 5 mM phosphate (pH 7.0), 2 mM EDTA, 75 mM e-aminocaproate, and dialyzed overnight against the same buffer. This preparation was kept at -90°C. The basal assay mixture for studying membrane phosphorylation contained, in a total volume of 0.2 ml, 200 pg membrane proteins (as determined by the Folin assay), 1O-20 PM [y-3 * P] ATP (1 to 4.1 O3 cmp/pmole; the Radiochemical Centre, Amersham, England), 50 mM Tris-HCl (pH 7.5), 1 mM aminophylline, Trasylol (Bayer, Leverkussen, Germany) 500 UIK/ml, and 10 mM MgS04. When specified, the medium was supplemented with 10 mM NaF, histones II A (Sigma, St. Louis, Missouri), 10e6 M cyclic or 5.10” M cyclic GMP. The phosphorylation was initiated by the addition of radioactive ATP and conducted at 30°C.
3. Results and discussion 2. Materials and methods Zymogen granules were isolated from rat pancreas as described previously [ 11. Secretory proteins were released in 0.2 M bicarbonate at pH 8.2. Membranes were purified by discontinuous gradient centrifugation according to Meldolesi [S] . They were utilized whitin the same day for the phosphorylation tests. 228
Fig. 1 shows that phosphorylation proceeded rapidly and that after 60 set a plateau was maintained for at least 5 min. This reaction was not limited by the hydrolysis of [r-32P] ATP by membranous ATPase. Indeed, the addition of an ATP-regenerating system made of 7.5 mM phosphoenolpyruvate and 25 U/ml pyruvate kinase did not improve the kinetics (data not shown). On the other hand, such a membrane preparatior North-Holland Publishing Company - Amsterdam
FEBS LETTERS
Volume 49, number 2 ,100
75
__.-. ____/---~
.:"f._._
0
___.__-.-.x~ "----,.-;k-~_____-______ _n
I
I
I
1
2
3
--------_____;
time
I
4
(min)
n
1 5
Fig. 1. Time-study of autophosphorylation of the membrane of zymogen granules with [T- ‘sP] ATP. This experiment was conducted as described under Materials and methods and is representative of 3 similar experiments. The phosphorylations were stopped by the addition of 1% Ccl, COOH and the pellets were filtered on Millipore AAWP (Bedford, U.S.A.). basal Results of one representative experiment (n=3). ( e) medium; (x-.x-.x) 10 mM NaF; ( o-c-o) l-lo6 M cyclic AMP.
had a moderate phosphoprotein phosphatase activity which was inhibited by 50% by 10 mM NaF (P. Robberecht, personal communication). Fig. 1 illustrates that in the presence of 10 mM NaF, protein phosphorylation exceeded control values. The plateau appears therefore to result from an equilibrium between the opposite activities of protein kinase and phosphoprotein phosphatase in the zymogen granule membrane. l.10e6 M cyclic AMP induced a moderate but consistent stimulation of autophosphorylation.
December 1974
Table 1 shows a comparison of the endogenous activity of membranes with a preparation of soluble protein kinase(s) partially purified from the rat pancreas. The autophosphorylation of membranes was only slightly stimulated by 1.10e6 M cyclic AMP or 5a104 M cyclic GMP and the protein kinase activity exerted by these membranes on added histones II A was unaffected by cyclic nucleotides. On the other hand, the activity of soluble protein kinase(s) was markedly stimulated by 1.10” M’cyclic AMP and 5.10’ M cyclic GMP. When this soluble preparation was added to membranes, the aggregate phosphorylation was lower than the sum of each autophosphorylation activities considered separately. The marked effect of cyclic nucleotides on this mixture suggests that the activity of soluble protein kinase(s) was predominating and contributed to the phosphorylation of membrane proteins. It can be hypothesized that a component of the crude soluble protein kinase preparation inhibited membrane protein kinase activity. It is more likely however that the affinity of semi-purified soluble protein kinase(s) for endogenous substrates was lower than that for membrane proteins. Competitive inhibitio might render the autophosphorylation of soluble protein negligible in the presence of zymogen granule membranes. As shown in Table 2, extraction of membranes with 0.5 M KC1 did not decrease the protein kinase activity measured on endogenous substrates or on added histones. In addition, this treatment did not affect the capacity of membranes to be phosphorylated in the
Table 1 Comparison of endogenous and exogenous protein kinase activities Proteins added
pmoles phosphate/mg
Soluble Histones no cyclic Zymogen nucleotide granule protein II A added membrane kinase (250 rg) (250 clg) (158 I.& + 0 0 47 0 + 0 21 + + 0 27 ___--__--___________-----------------___---___ + 0 + 125 0 + + 531
protein. 10 min
cyclic AMP
cyclic GMP
1.1O-6 M 58 33 59
5.10-6 M 52 26 44
125 1004
116 811
Assays were performed in the presence of 10 mM NaF. Results of one representative experiment (n=4).
229
FEBS LETTERS
Volume 49, number 2
Table 2 Protein kinase activity of the membrane of zymogen granules after treatment with 0.5 M KC1 Proteins added
No addition Soluble protein kinase (250 pg) Histones II A (250 fig)
pmoles phosphate/mg protein. 5 min Control membranes
Washed membranes
48 110 64
73 129 66
Purified membrane pellets were suspended at 0°C in 50 mM Tris-HCl (pH 7.5) and Trasylol500 UIK/ml in the presence or absence (control) of 0.5 M KCl. These suspensions were recentrifuged for 10 min at 2500 X g. The recovered membranes were then utilized for phosphorylation tests in the presence of 10 mM NaF. Results of one respresentative experiment (n=3).
cm
December 1974
presence of exogenous protein kmase(s). Thus it appears that protein kinase activity and protein substrates for this enzyme were intrinsic parts of the zymogen membrane and can not be assimilated to exogenous proteins simply contaminating the membrane. Individual’protein components of the zymogen granule membrane were analyzed by polyacrylamide gel electrophoresis in sodium dodecylsulfate. Staining with Coomassie blue revealed at least 26 distinct bands (fig. 2A). The crude soluble protein kinase preparation when submitted to similar electrophoresis showed 10 to 12 bands, the main component having a mol.wt of 66 000 (fig. 2B). As can be seen in fig. 3,8 to 10 components of
moved
Fig. 2. Polyacrylamide gel electrophoresis in 0.1% sodium dodecylsulfate after 10 min incubation of: A: zymogen granule membrane from rat pancreas (a indicates the protein band most markedly phosphorylated in the high molecular weight range, as illustrated in fig. 3). B: semi-purified soluble protein kinase preparation from rat pancreas. C: a mixture of the two preceding extracts. The standard incubation medium described under Materials and methods was suplemented with 10 mM NaF. Each incubation was stopped by the addition of 0.1 ml 0.03 M phosphate buffer (pH 7.2), 3% (v/v) fl-mercaptoethanol, 24% sucrose, 6% sodium dodecylsulfate. These samples were heated for 2 min at 100°C. 25 ~1 samples were then applied to polyacrylamide gels containing 10% acrylamide, 0.1% sodium dodecylsulfate, 0.1 M sodium phosphate (pH 7.2), 0.125% persulfate, and 0.0135% Temed (from Fluka, Buchs, Switzerland). Electrophoresis was carried out for 270 min at 8 mA/tube. These gels were stained with Coomassie blue in 10% acetic acid-25% isopropanol and destained in 10% acetic acid-lo% isopropanol.
230
December 1974
FEBS LETTERS
Volume 49, number 2 c
w llnOUT 600
AMP
w Ill4
ACTIVITY
BASAL
CAMP
lq6 M
BASAL
ACTIVITY
500
coo
l-l
-100
300 U300 ” _ 100 v) 0
Iso
E
-40
0
-30 -20
u 3
-10
s
I* Y 0
3 ” w -8 600
ii-
0 *
0 s
KINASE
PROTEIN
K INASE
PROTEIN
>
I
ADDED
D
500
Y I-
4
z ;
p 400 0.
100
CJ300 200
loo
\
-\
0_1
0
I
IO
20
SLICE
I
30
‘\ I
LO
!
50
No
r
0
,
I
to
20 SLICE
I
LO
30
I
z
I
:8
0 0
30 20
2
IO
50
No
Fig. 3. Distribution of aZP among the proteins of phosphorylated zymogen granule membranes, after separation by polyacrylamide gel electrophoresis in 0.1% sodium dodecylsulfate. After 10 min incubation carried out under the same general conditions as those in fig. 2, in the presence of 10 mM NaF and with or without 1=106 M cyclic AMP, 100 nl of SDS-solubilized proteins were applied to each gel. After electrophoresis, the gels were fixed overnight in 10% acetic acid-25% isopropanol, washed in 10% acetic acid-10% isopropanol and cut into 2 mm slices. The radioactivity was.measured by scintillation counting. The autophosphorylation profile of the isolated crude soluble protein kinase preparation is illustrated by broken lines in the two lower panels. The linear relationship between molecular weight and the distance of migration was standardized with bovine serum albumin, hen ovalbumin, bovine carboxypeptidase A, bovine ribonuclease and horse heart cytochrome c.
the membrane with mol.wt ranging from 12 000 to 70 000 were phosphorylated by endogenous kinase. There was no specific effect of cyclic AMP on a particular protein species. These results are at variance with those reported by MacDonald and Ronzio [6] while our work was in progress. These authors noticed that more than 70% of 32P was incorporated into a single protein of approximate mol.wt 130 000 in zymogen granule membranes incubated for 5 set
without aminophylline and NaF (this component was isolated by SDS polycrylamide gel electrophoresis at pH 2.4). After incubation in the presence of the exogenous kinase, we detected 8 phosphorylated components. Cyclic AMP increased the incorporation of y-32P in all molecular species but its effect was particularly pronounced for a protein with a mol.wt of approx. 70 000 (Band a in fig. 2A). In conclusion, it appears that 8 to 10 proteins in 231
Volume 49, number 2
FEBS LETTERS
the zymogen granule membrane can be phosphorylated by an endogenous protein kinase. In addition, soluble protein kinase(s) can also phosphorylate the membrane of zymogen granules. The endogenous membrane protein kinase activity was only slightly dependent on cyclic nucleotides. If we assume that this kinase remained unaltered during membrane isolation, this observation might suggest that the in vivo phosphorylation of zymogen granule membrane is not under hormonal control. This conclusion would contradict our previous hypothesis, based on intact cell preparation [l] , according to which phosphorylation of specific proteins could change granule membrane properties and facilitate the fusion with the plasma membrane during exocytosis. It is however possible to reconcile the data obtained on pancreas fragments and pancreas extracts if we admit that rises in the intracellular levels of cyclic GMP or cyclic AMP entail a dissociation of cytoplasmic protein kinase(s) am’ that the liberated catalytic subunit(s) are secondarily transferred on zymogen granule membranes, resulting in ‘in situ’ phosphorylation of 8 to 10 constituent proteins. This hypothesis finds some support __ in recent observations. Korenman et al. [7] demonstrated a translocation of cytoplasmic protein kinase towards microsomes, in a cyclic AMP-independent form, after stimulating rat uterus with isoproterenol. Our own preliminary results (unpublished data) indicate an important rise in membrane bound protein kinase activity when zymogen granule membranes of
232
December 1974
rat pancreas were incubated at O’C in the presence of semi-purified soluble protein kinase and of 1 *lo” M cyclic AMP.
Acknowledgements This work was supported by Fonds de la Recherche Scientifique MBdicale (Belgium), grant No. 20 403. We wish to express our thanks to Dr Wald (Bayer, Brussels), for the gift of Trasylol and to Mrs Ballinckx and Mr Gelston for preparation of the manuscript.
References [l] Lambert, M., Camus, I. and Christophe, J. (1973) Biochem.
Biophys. Res. Commun. 52,935-942. [2] Case, R. M., Johnson, M., Scratcherd, T. and Sherratt, H. S. A (1972) J. Physiol. 223,669-684. [31 Rutten, W. J., de Pont, J. J. H. H. M. and Bonting, S. L. (1972) Biochim. Biophys. Acta 274, 201-213. [41 Robberecht, P., Deschodt-Lanckman, M., de Neef, P., Borgeat, P. and Christophe, J. (1974) FEBS Lett. 43. 139-143. ISI Meldolesi, J., Jamieson, J. D., Palade, G. E. and Evans, L. M (1971) J. Cell. Biol. 49, 109-129. I61 MacDonald, R. J. and Ronzio, R. A. (1974) FEBS Lett. 40,203-206. [71 Korenman, S. G., Bhalla, R. C., Sanborn, B. M. and Stevens, R. H. (1974) Science 183,430-432.